Power grid transmission section identification method based on improved PageRank algorithm and network conversion

Abstract

The application discloses a power grid power transmission section identification method based on an improved PageRank algorithm and network conversion, first, an improved PageRank algorithm is used to calculate a node criticality index, and a network constraint coefficient is calculated according to a network topological structure and is used to correct a restart vector; second, a key power transmission section search topological graph is constructed through network conversion and node criticality index calculation, and identification of the key power transmission section is converted into search of a shortest path; finally, according to energy transaction between a generator node and a load node, a search range of the key power transmission section is set, a related cut set with the highest criticality is found, and through checking and correction, an identification result of the key power transmission section is obtained. The application can well improve the deficiency of the original PageRank algorithm when applied to the power grid, can accurately identify the power transmission section of the power grid, provides a key monitoring guide for power grid monitoring and dispatch personnel, and better guarantees the safe and stable operation of the power grid.

Description

Technical Field

[0001] This invention belongs to the technical field of power system transmission section identification methods, and specifically relates to a power grid transmission section identification method based on an improved PageRank algorithm and network transformation. Background Technology

[0002] With the continuous development of power systems, their topologies and operating states are becoming increasingly complex, posing growing challenges to their safety and stability. Transmission lines within a power system play a decisive role in the safe operation of the grid. If a transmission line in a transmission section goes out of service due to a fault, it can potentially cause drastic changes in the power flow of other transmission lines within the cut-off set, even triggering serious faults and leading to large-scale power outages. Therefore, accurately and effectively identifying transmission lines is of great significance for the safe and stable operation of the power grid and for preventing large-scale power outages.

[0003] The PageRank algorithm is a ranking algorithm for calculating the criticality of web pages. In recent years, some scholars have drawn analogies between power systems and the Internet, combining the algorithm with the physical characteristics of power systems to identify critical nodes within them. The original PageRank algorithm posits that if web page i is linked from web page j, then web page i gains some of the importance of web page j, thus increasing its own importance. This importance is defined as the PageRank value of web page i, denoted as PR(i). The original PageRank algorithm is highly accurate and effective in calculating the criticality of web pages. However, due to the structural differences between power systems and the Internet, directly using the original PageRank algorithm to calculate the criticality of nodes in power systems presents the following problems:

[0004] In research aimed at improving the PageRank algorithm, most studies only modify the Google matrix, while the restart vector often uses a unit vector. This indicates that all nodes in the network have the same restart probability, which clearly does not reflect the actual physical conditions of the power grid. Furthermore, most methods require node classification before computation, but the node states in a real power system are constantly changing. Therefore, node classification increases computational complexity and reduces the versatility of these algorithms. Existing transmission section identification methods also suffer from problems such as unreasonable partitioning leading to reduced accuracy, omission of transmission sections within sub-regions, and insufficient consideration of the actual operating state of the power grid. Summary of the Invention

[0005] To address the aforementioned problems, the present invention aims to provide a power grid transmission section identification method based on an improved PageRank algorithm and network transformation. This method effectively overcomes the shortcomings of the original PageRank algorithm when applied to power grids, accurately identifying power grid transmission sections and providing key monitoring guidance for power grid monitoring and dispatching personnel, thereby better ensuring the safe and stable operation of the power grid. The specific details are as follows.

[0006] A method for identifying power grid transmission sections based on an improved PageRank algorithm and network transformation includes the following steps:

[0007] Step 1: Construct an optimal operation model of the power system and solve for the energy trade volume P between generator nodes and load nodes. g,d,t ;

[0008] Step 2: Based on the energy transaction volume P between the generator node and the load node g,d,t Set up a virtual circulation and correct the Google matrix G in the original PageRank algorithm;

[0009] Step 3: Calculate the corrected restart vector e* of the node network constraint coefficients C(i) to obtain the improved PageRank algorithm formula;

[0010] Step 4: Set the keyness PR for all nodes 0 Initialize the values ​​and set the criticality PR of the nodes. 0 The initial value vector is substituted into the improved PageRank algorithm formula;

[0011] Step 5: Iteratively calculate the node criticality until the absolute value of the difference between the (k+1)th result and the kth result is less than ε, then output the node criticality PR. k ;

[0012] Step 6: Based on node criticality PR k A new network topology map is constructed through network transformation for the identification of power transmission sections;

[0013] Step 7: Based on P g,d,t Determine the area for searching transmission sections, and find the shortest path within the corresponding area using a shortest path search algorithm;

[0014] Step 8: Restore the shortest path to a set of transmission lines. If this set of lines can form a cut set, proceed to step 11; otherwise, proceed to step 9.

[0015] Step 9: Modify the current transmission line set by adding or replacing other lines related to the shortest path; if a cut set is formed after modification, proceed to step 11; otherwise, proceed to step 10.

[0016] Step 10: The current generator node-load node pair cannot form a transmission section. Replace with other generator node-load node pairs and return to step 7.

[0017] Step 11: If the power flow direction of the lines within the cut set is consistent, proceed to step 12; otherwise, return to step 10.

[0018] Step 12: Output the transmission section identification results and end the program.

[0019] Furthermore, in step 1, based on the fundamental principle of power flow transmission distribution factor, the constructed power system optimization operation model includes the energy trading volume P between generator nodes and load nodes. g,d,t The relevant constraints are solved using equations (1)-(4) to obtain the power flow components of the transmission line:

[0020] The output of generator g is equal to the sum of all transactions between the load and generator node g:

[0021]

[0022] The power of load node d is equal to the sum of the transaction volumes between all generators and load d:

[0023]

[0024] The component of energy exchange between all generator node-load node pairs on line l The sum equals the actual power flow P of the line. l,t :

[0025]

[0026] Based on the energy exchange volume and power transmission distribution factor between generator node g and load node d, calculate the power flow components on line l.

[0027]

[0028] Where g and d represent generator nodes and load nodes, respectively; F, D, and T represent the set of generator nodes, the set of load nodes, and the set of scheduling time periods, respectively; L is the set of lines; P g,d,t P represents the energy exchange between generator node g and load node d at time t. g,t P represents the active power at generator node g at time t. d,t P represents the active power of load node d at time t. l,t The actual power flow of transmission line l at time t; x represents the energy exchange between generator node g and load node d at time t on transmission line l.ij X is the impedance value of branch l. ig This represents the element in the i-th row and g-th column of the power grid impedance matrix X. jg X id X jd All of these are elements in the power grid impedance matrix X.

[0029] Furthermore, step 2 is based on the energy transaction volume P between the generator node and the load node. g,d,t The virtual circulation is set up and the Google matrix G in the original PageRank algorithm is corrected, as follows:

[0030] Step 2.1: Add a virtual power flow pointer to g from load node d, with the power flow size being the energy trade volume P between generator node and load node. g,d,t ;

[0031] Step 2.2: Set up virtual current conversion: Assume that the generator sets on a certain generator node transmit active power flow to multiple load nodes through transmission lines; based on the energy transaction data between generator nodes and load nodes calculated by the power system optimization operation model, the active power transmission path between generator nodes and load nodes is redirected from the load nodes to the generator node, and the weight of the virtual circulation is set according to the specific transaction volume. Then, all active power sent out by the generator returns to the generator node in the form of virtual circulation.

[0032] By setting up a virtual circulation, the Google matrix G in the PageRank calculation formula is modified. The modified Google matrix is ​​denoted as G. * The specific formula is as follows:

[0033]

[0034]

[0035] In equation (6), P ij For the active power flow of transmission line ij, P out (j) represents the outflow power of node j when considering virtual circulation.

[0036] Furthermore, step 3 specifically includes:

[0037] Step 3.1: Correct the restart vector e in the PageRank algorithm, denoted as e* after correction, as follows:

[0038]

[0039] Where C(i) is the network constraint coefficient of node i, which is determined by the topology of the power grid; n is the number of nodes in the power grid;

[0040] Step 3.2: Calculate the network constraint coefficients C(i) for node i:

[0041]

[0042]

[0043]

[0044]

[0045]

[0046] Where q represents the common neighbor nodes of nodes i and j, and p ij p represents the proportion of node j to all adjacent nodes of node i. iq p represents the proportion of node q to all adjacent nodes of node i. qj Let Γ(i) represent the proportion of node j to all adjacent nodes of node q; let Γ(i) be the set of adjacent nodes of node i; let B(j) and B(u) be the three-level degrees of node j and node u, respectively; let B(i) and k(i) be the degrees of node i considering three-level and single-level neighbor nodes, respectively; x ij Let be the reactance value of line l; w be the neighboring nodes of node i, v be the neighboring nodes of node w; Γ(w) be the set of adjacent nodes of node w; k(v) be the degree of a single layer of node v; a ij Let be the connection weight between nodes i and j; N is the set of all nodes in the network.

[0047] Step 3.3: The improved PageRank algorithm calculation formula is as follows:

[0048] PR k+1 =δ·G * ·PR k +(1-δ)e * (13)

[0049] Where δ is the drag coefficient, PR k+1 and PR k These are the node criticality values ​​calculated in the (k+1)th and kth iterations, respectively.

[0050] Furthermore, the specific method for network conversion in step 6 is as follows: nodes and edges in the power system topology graph are interchanged, that is, transmission lines and buses are mutually converted to obtain a new topology graph. Simultaneously, the criticality index (PR) of the nodes in the original topology graph is adjusted. k The reciprocal of the result is used as the converted line weight.

[0051] Furthermore, step 7 specifically includes:

[0052] Step 7.1: Based on the energy trading volume between nodes, determine the range between generator nodes and load nodes where the energy trading volume is greater than the set threshold as the range for the transmission section search;

[0053] Step 7.2: Within the search range of the transmission section, obtain the shortest path using the shortest path search algorithm.

[0054] The beneficial effects of this invention are:

[0055] (1) This invention can accurately identify power grid transmission sections by improving the PageRank algorithm and network transformation.

[0056] (2) The present invention calculates the network constraint coefficients based on the actual topology of the power grid and corrects the restart vector in the PageRank algorithm, thereby avoiding the situation where the reconnection probability of nodes is equal during the iteration process; at the same time, it does not require the classification of nodes, which can improve the calculation speed and enhance the versatility of the PageRank algorithm.

[0057] (3) This invention constructs a transmission section identification topology map through network transformation. Then, the search range is determined based on the transaction volume between generator nodes and load nodes, and the relevant line set is found by searching for the shortest path. After correcting and verifying the line set, the transmission section identification result is obtained. This eliminates the need for network partitioning to identify transmission sections, thus avoiding problems such as reduced accuracy due to unreasonable partitioning and omission of transmission sections within sub-regions, thereby improving the accuracy and practicality of the identification. Attached Figure Description

[0058] Figure 1 This is a flowchart of the steps of the power grid transmission section identification method based on the improved PageRank algorithm and network conversion of the present invention.

[0059] Figure 2 This is a schematic diagram of the virtual circulation of the present invention.

[0060] Figure 3 This is a schematic diagram of the network conversion described in this invention.

[0061] Figure 4 This is a schematic diagram of the power transmission section identification described in this invention.

[0062] Figure 5 This is a bar chart showing the node criticality in the example described in this invention.

[0063] Figure 6 This is a schematic diagram of the power transmission section 1 in the example described in this invention.

[0064] Figure 7 This is a bar chart of node criticality obtained by the virtual node method in the example described in this invention. Detailed Implementation

[0065] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0066] This invention presents a power grid transmission section identification method based on an improved PageRank algorithm and network transformation. The method comprises steps including calculating the energy transaction volume between generator nodes and load nodes, correcting the restart vector e, iteratively calculating the node criticality degree, network transformation, and transmission section identification. First, based on PTDF and real-time power flow data of the power grid, the active power transaction between generator nodes and load nodes is calculated to accurately grasp the real-time state of the power grid. Second, the improved PageRank algorithm is used to calculate the node criticality index. A virtual circulating current is set according to the energy transaction volume between generator nodes and load nodes to correct the Google matrix. Simultaneously, network constraint coefficients are calculated considering the network topology to correct the restart vector e, thereby avoiding situations where restart probabilities are equal. Then, through network transformation and node criticality index calculation, a transmission section search topology graph is constructed, transforming transmission section identification into a shortest path search. This avoids problems such as reduced accuracy due to unreasonable partitioning and the omission of transmission sections within sub-regions. Finally, based on the energy trading situation between generator nodes and load nodes, the search range of transmission sections is set, the relevant cut sets with the highest criticality are found, and the identification results of transmission sections are obtained through verification and correction.

[0067] The flowchart of the power grid transmission section identification method based on the improved PageRank algorithm and network transformation of this invention is as follows: Figure 1 As shown, the specific steps include:

[0068] Step 1: Based on the power system optimization operation model, solve for the energy transaction volume P between generator nodes and load nodes. g,d,t .

[0069] Based on the fundamental principle of power flow transmission distribution factor (PTDF), the energy trade volume P between generator nodes and load nodes is included in the constructed power system optimal operation model. g,d,t The relevant constraints can be solved by equations (1)-(4) to obtain the power flow components of the transmission line.

[0070] Equation (1) indicates that the output of generator g is equal to the sum of the transactions between all loads and generator node g; Equation (2) indicates that the power of load node d is equal to the sum of the transactions between all generators and load d; Equation (3) indicates the component of energy transactions between all generator node-load node pairs on line l. The sum equals the actual power flow P of the line. l,t Equation (4) calculates the power flow components on line l based on the energy exchange between generator node g and load node d and the PTDF.

[0071]

[0072]

[0073]

[0074]

[0075] Where g and d represent generator nodes and load nodes, respectively; F, D, and T represent the set of generator nodes, the set of load nodes, and the set of scheduling time periods, respectively; P g,d,t P represents the energy exchange between generator node g and load node d at time t. g,t P represents the active power at generator node g at time t. d,t P represents the active power of load node d at time t. l,t The actual power flow of transmission line l at time t; Let x be the component of the energy exchange between generator node g and load node d at time t on transmission line l. ij X is the impedance value of branch l. ig This represents the element in the i-th row and g-th column of the power grid impedance matrix X. jg X id X jd similar.

[0076] Step 2: Based on the energy transaction volume P between the generator node and the load node g,d,t We set up a virtual circulation and corrected the Google matrix G in the original PageRank algorithm.

[0077] The specific improvement measures are as follows:

[0078] Step 2.1: Add a virtual power flow from load node d to generator node g, with the power flow size being the energy exchange volume P between generator node and load node. g,d,t .

[0079] Step 2.2: Set up virtual circulation

[0080] The virtual circulation settings are as follows Figure 2 As shown. Figure 2 In this system, the generator units at node 1 transmit active power flow to nodes 3, 6, 7, 10, and 11 via transmission lines. Based on the energy trading data between generator nodes and load nodes calculated using the power system optimization operation model, the active power transmission path between generator nodes and load nodes can be redirected from the load nodes to the generator nodes, and the weight of the virtual circulating current can be set according to the specific trading volume. For example... Figure 2As shown by the dashed line, all active power delivered by the generator returns to the generator node in the form of a virtual circulation. Without adding virtual nodes, the load node's out-degree is no longer equal to 0, thus avoiding the black hole effect while increasing the influence of both the generator and load nodes.

[0081] By setting up a virtual circulation, the Google matrix G in the PageRank calculation formula is modified. The modified matrix is ​​denoted as G. * The specific formula is as follows:

[0082]

[0083]

[0084] In equation (6), P ij For the active power flow of transmission line ij, P out (j) represents the outflow power of node j when considering virtual circulation.

[0085] Step 3: Calculate the node network constraint coefficients and correct the restart vector e.

[0086] Specifically as follows:

[0087] Step 3.1: This invention modifies the restart vector e in the PageRank algorithm, denoted as e*, thereby solving the problem that all nodes have equal relink probabilities, as detailed below.

[0088]

[0089] Where C(i) is the network constraint coefficient of node i, which is determined by the topology of the power grid; n is the number of nodes in the power grid.

[0090] Step 3.2: Calculate the network constraint coefficients C(i) for node i:

[0091]

[0092]

[0093]

[0094]

[0095]

[0096] Where q represents the common neighbor nodes of nodes i and j, and p ij p represents the proportion of node j to all adjacent nodes of node i. iq p represents the proportion of node q to all adjacent nodes of node i. qjLet Γ(i) represent the proportion of node j to all adjacent nodes of node q; let Γ(i) be the set of adjacent nodes of node i; let B(j) and B(u) be the three-level degrees of node j and node u, respectively; let B(i) and k(i) be the degrees of node i considering three-level and single-level neighbor nodes, respectively; x ij Let be the reactance value of line l; w be the neighboring nodes of node i, v be the neighboring nodes of node w; Γ(w) be the set of adjacent nodes of node w; k(v) be the degree of a single layer of node v; a ij Let be the connection weight between nodes i and j; N is the set of all nodes in the network.

[0097] According to the structural hole theory, the smaller the network constraint coefficient C(i) of a node, the more neighboring nodes node i has and the closer the connections with them, thus forming more structural holes. Nodes with more structural holes are more critical in the energy transfer process.

[0098] In addition, Equation (12) introduces the reciprocal of the transmission line reactance as the weight for calculating the connection relationship between nodes. The smaller the reactance between nodes, the closer their electrical distance and the tighter the connection.

[0099] In summary, the improved network constraint coefficient C(i) takes into account the topological characteristics of node i itself and its outer neighbors, and assigns weights to the lines between nodes, which can more accurately reflect the importance of node i.

[0100] The improved PageRank algorithm calculation formula is as follows:

[0101] PR k+1 =δ·G * ·PR k +(1-δ)e * (13)

[0102] Where δ is the drag coefficient, which is usually taken as δ = 0.85 based on experience.

[0103] Step 4: Set the keyness PR for all nodes 0 Initialize the values ​​and set the criticality PR of the nodes. 0 The initial value vector is substituted into the improved PageRank algorithm formula.

[0104] Step 5: Iteratively calculate the node criticality until the absolute value of the difference between the (k+1)th result and the kth result is less than ε, then output the node criticality PR. k .

[0105] Step 6: Based on node criticality PR k A new network topology map is constructed through network transformation for the identification of power transmission sections.

[0106] Based on the node criticality obtained in step 6 and the energy transaction volume between generator nodes and load nodes obtained in step 1, this invention proposes a new method to construct a transmission section search topology map using the relevant data obtained above, so as to accurately and effectively identify the transmission section.

[0107] Based on the power system topology and node criticality indices, a transmission section identification topology map is constructed through network transformation. The specific method of network transformation involves swapping nodes and edges, corresponding to the mutual transformation of transmission lines and buses in the power system. Simultaneously, the node criticality index is used as the weight of the transformed line. The transformation process is as follows: Figure 3 As shown.

[0108] Figure 3 The left side depicts a simple power system topology, with nodes 1-5 interconnected by lines a, e, and e. By transforming the nodes, a new topology is obtained. Figure 3 (Right side), where the weight of the edge is the reciprocal of the PR value of the node in the original topology network. Figure 3 The edge between nodes a and b on the right side has a weight of 1 / PR(2).

[0109] Step 7: Based on P g,d,t The area to be searched for transmission sections is determined, and the shortest path within the corresponding area is found using a shortest path search algorithm.

[0110] Based on energy trading data between nodes, the search area for transmission sections is determined. These generator nodes and load nodes have larger energy transactions, and the corresponding energy transmission paths are more critical to the power system. If related lines go out of service due to faults, it will significantly impact the power supply to the loads and could potentially trigger power flow shifts, causing faults on other lines. Therefore, identifying transmission sections between generator nodes and load nodes with large energy transactions is both necessary and reasonable. Specific identification methods are as follows... Figure 4 As shown.

[0111] like Figure 4 The greater the energy exchange between nodes i and j, the more important the energy transmission is on the related lines along their transmission paths. Therefore, the search range for transmission sections is determined to be the region between i and j. After network topology transformation, a new network topology is obtained. The shortest path is then obtained using the shortest path search algorithm, corresponding to... Figure 4 This is a set of routes. Therefore, searching for routes is transformed into searching for nodes.

[0112] Meanwhile, the shortest path is determined based on the PR values ​​of common nodes in the transmission lines. The shorter the path distance, the higher the PR value of the nodes in the resulting set of transmission lines, indicating that the path passes through nodes with higher criticality, and thus the higher the criticality of the resulting set of transmission lines. For example, if the criticality indices of nodes e and f are greater than those of nodes a and b, then the criticality of transmission section 1 is higher than that of transmission section 2. According to the method proposed in this invention, the shortest path found in the newly constructed network topology graph corresponds to the set of transmission lines with the highest node criticality in the original network, namely the most vulnerable energy transmission link between ij. This set of transmission lines is then corrected and verified to identify the most critical set of transmission sections between ij.

[0113] Step 8: Restore the shortest path to a set of transmission lines. If this set of lines can form a cut set, proceed to step 11; otherwise, proceed to step 9.

[0114] Step 9: Modify the current set of transmission lines by adding or replacing other lines related to the shortest path; if a cut set is formed after modification, proceed to step 11; otherwise, proceed to step 10.

[0115] Step 10: The current generator node-load node pair cannot form a transmission section. Replace with other generator node-load node pairs and return to step 7.

[0116] Step 11: If the power flow direction of the lines within the cut set is consistent, proceed to step 12; otherwise, return to step 10.

[0117] Step 12: Output the transmission section identification results and end the program.

[0118] Example

[0119] (1) Example Introduction

[0120] To verify the effectiveness and applicability of the method proposed in this invention, the IEEE 118-bus system was used as a test case to identify the transmission section.

[0121] (2) Analysis of the results of the example

[0122] By calculating the transaction volume between generator node-load node pairs and the network constraint coefficient, an improved PageRank algorithm is used to obtain the criticality index of the IEEE 118-node system, such as... Figure 5As shown in the diagram, nodes 49, 24, 70, and 100 have high PR values ​​and are among the most critical nodes in the current operating state. Combining generator output data, power flow distribution, and network topology, it can be seen that these nodes are important power supply nodes or located at network hubs. For example, node 49 connects 12 transmission lines, making it a crucial network topology connection point and also an important power supply node in the system. Nodes 24 and 70 have generator active powers of 600MW and 400MW respectively, supplying power to many important nodes. Although the generator output of node 100 is not large, it is located at the network topology hub, connecting nodes 82-99 on the left and nodes 103-112 on the right, such as... Figure 6 As shown. If this node fails and goes out of service, it will cause the right subnetwork to completely disconnect from the left main network, resulting in a drastic change in power distribution. Therefore, during cutset identification, the lines related to node 100 will have a higher probability of forming a transmission section. Thus, by combining the analysis of the node's own attributes and its network topology location, it is demonstrated that the improved PageRank algorithm of this invention can accurately calculate the node criticality index, providing a data foundation for subsequent transmission section identification.

[0123] The energy transaction volumes between generator nodes and load nodes are sorted in descending order, and the top 10 generator node-load node pairs with the largest transaction volumes are selected as the search range for transmission sections. The transmission sections are identified using the method proposed in this invention, and the results are shown in Table 1 below.

[0124] Table 1. Transmission section identification results under the IEEE 118-bus system

[0125]

[0126]

[0127] By calculating node criticality indices and network transformations, the method proposed in this invention was used to identify transmission sections for 10 pairs of generator-load nodes, resulting in 7 sets of transmission sections. (See Table 1 and...) Figure 6It is known that the lines of transmission sections 1 and 2 are both composed of lines 100-103, 100-104, and 100-106. Topology analysis reveals that the power transmission paths of generator node 92 and nodes 108 and 105 are highly similar, and all pass through the aforementioned three lines. Therefore, it is reasonable that cut sets 1 and 2 have the same constituent lines. Node 100 is the pivotal location connecting generator node 92 with load nodes 105 and 108, and all three lines of cut sets 1 and 2 are connected to node 100, indicating its high criticality. This also verifies the high accuracy of the improved PageRank algorithm. After disconnecting line 100-103, it was found that its original 322MW active power flow was entirely transferred to the other two lines. Among them, line 100-104 increased from 147MW to 329MW, an increase of 124%, far exceeding its maximum power flow limit of 175MW. Lines 100-106 increased from 155MW to 295MW, a 90% increase, and also experienced power flow exceeding limits. This demonstrates the high interconnectivity of lines in cut sets 1 and 2; if one line were to become out of service, the power flow on the other two lines would surge. Power flow exceeding limits caused all lines in the cut set to disconnect, resulting in nodes such as 103-112 becoming completely isolated from the main grid and triggering a large-scale power outage.

[0128] From a topological perspective, the four lines constituting cutset 3 divide the entire system into two roughly equal subsystems. It also separates generator node 24 from load node 66. Disconnecting lines 23-24 causes a reversal in power flow on lines 65-68 and 49-69, with power flow changes of 226MW and 32MW before and after the disconnection, respectively. Simultaneously, the power flow on line 47-69 increases by 33MW. This demonstrates that the power flow transfer between lines in cutset 3 is very significant, easily causing power flow fluctuations and fault propagation, requiring close monitoring. Cutsets 5, 6, 7, and 8 also connect generator nodes and load nodes, or divide the power grid into different sub-regions. Several lines constituting these cutsets are responsible for transmitting active power from the sub-region where the generator is located to the sub-region where the load is located. If a line within a cutset goes out of service due to a fault, it will cause significant changes in the power flow on other lines, and may even lead to further expansion of the fault.

[0129] To further compare and analyze, this invention uses the power flow data from step 1 as a basis and employs the virtual node method to calculate the node criticality index. The results are as follows: Figure 7 As shown in Table 2, the GN partitioning method was then used to identify the transmission sections.

[0130] Table 2. Transmission section identification results of the GN partitioning method for the IEEE 118-bus system

[0131]

[0132] Depend on Figure 7 It can be seen that nodes 100, 55, 34, 19, and 15 have high criticality indices. These nodes are mostly important load nodes or occupy critical positions in the topology. For example, node 100 is located between generator node 92 and load node 105, and connects to eight transmission lines, thus making it a critical node in the system. This is consistent with the calculation results of the method proposed in this invention. However, some important generator nodes with high output, such as nodes 49, 24, and 92, have very low criticality indices calculated using the virtual node method. This does not match the actual situation, indicating that the virtual node method does not fully consider the importance of generator nodes. Furthermore, nodes 42, 44, and 45 have high criticality indices, but in reality, these nodes have low loads and are located at the network edge, not being critical nodes in the system. Therefore, the PageRank algorithm improved by adding virtual nodes has certain errors. In contrast, the method proposed in this invention has higher accuracy and better reflects the actual situation.

[0133] Table 2 shows that the GN partitioning method identifies four sets of transmission sections. Cut set 1 is the same as the cut set 1 in the method proposed in this invention. The three lines constituting cut set 1 can isolate ten nodes, such as 103-112, from the main grid, as shown below. Figure 6 As shown, this creates an active power imbalance, thus representing a transmission section of the system. Cut set 2, identified by the GN partitioning method, consists of lines 77-82, 82-96, etc., dividing the network into two independently operating sub-regions. Similarly, disconnecting line 99-100, which has the largest power flow in cut set 2, increases the power of line 77-82 from 72MW to 105MW, only reaching half of the line capacity. The other four lines only increase by an average of 26MW, with the increased power flow less than 1 / 3 of the line capacity limit. In comparison, the cut set identified by the method proposed in this invention exhibits a more pronounced power flow shift phenomenon and poses a greater threat to the system. Cut set 3, identified by the GN partitioning method, has a high degree of overlap with cut set 2, representing the common edge of the two sub-regions after GN partitioning. Cut set 4 is similar to cut set 5 identified by the method proposed in this invention, but the latter consists of fewer lines, and the nodes connected by the cut set are more critical. Therefore, cut set 5 identified by the method proposed in this invention is more reasonable and accurate. Furthermore, the method proposed in this invention also identifies cut sets 3, 6, and 7 located within sub-regions of the power grid, thus overcoming the shortcomings of the GN partitioning method. In summary, the comparative analysis shows that the transmission sections identified by the method proposed in this invention are more accurate and reasonable, with a lower possibility of omissions. The above analysis and testing demonstrate that the transmission sections identified by the method proposed in this invention are correct and reasonable.

[0134] The above description is merely a specific embodiment of the present invention, but it does not limit the scope of patent protection of the present invention. Any equivalent changes or substitutions made using the content of the present invention specification and drawings, and any direct or indirect application to other related technical fields, should be included within the scope of protection of the present invention.

Claims

1. A method for identifying power grid transmission sections based on an improved PageRank algorithm and network transformation, characterized in that, Includes the following steps: Step 1: Construct an optimal operation model of the power system and solve for the energy trade volume P between generator nodes and load nodes. g,d,t ; Step 2: Based on the energy transaction volume P between the generator node and the load node g,d,t Set up a virtual circulation and correct the Google matrix G in the original PageRank algorithm; Step 3: Calculate the corrected restart vector e* of the node network constraint coefficients C(i) to obtain the improved PageRank algorithm formula; Step 4: Set the keyness PR for all nodes 0 Initialize the values ​​and set the criticality PR of the nodes. 0 The initial value vector is substituted into the improved PageRank algorithm formula; Step 5: Iteratively calculate the node criticality until the absolute value of the difference between the (k+1)th result and the kth result is less than ε, then output the node criticality PR. k ; Step 6: Based on node criticality PR k A new network topology map is constructed through network transformation for the identification of power transmission sections; Step 7: Based on P g,d,t Determine the area for searching transmission sections, and find the shortest path within the corresponding area using a shortest path search algorithm; Step 8: Restore the shortest path to a set of transmission lines. If this set of lines can form a cut set, proceed to step 11; otherwise, proceed to step 9. Step 9: Modify the current transmission line set by adding or replacing other lines related to the shortest path; if a cut set is formed after modification, proceed to step 11; otherwise, proceed to step 10. Step 10: The current generator node-load node pair cannot form a transmission section. Replace with other generator node-load node pairs and return to step 7. Step 11: If the power flow direction of the lines within the cut set is consistent, proceed to step 12; otherwise, return to step 10. Step 12: Output the transmission section identification results and end the program.

2. The power grid transmission section identification method based on the improved PageRank algorithm and network transformation according to claim 1, characterized in that, In step 1, based on the fundamental principle of power flow transmission distribution factor, the constructed power system optimization operation model includes the energy transaction volume P between generator nodes and load nodes. g,d,t The relevant constraints are solved using equations (1)-(4) to obtain the power flow components of the transmission line: The output of generator g is equal to the sum of all transactions between the load and generator node g: The power of load node d is equal to the sum of the transaction volumes between all generators and load d: The component of energy exchange between all generator node-load node pairs on line l The sum equals the actual power flow P of the line. l,t : Based on the energy exchange volume and power transmission distribution factor between generator node g and load node d, calculate the power flow components on line l. Where g and d represent generator nodes and load nodes, respectively; F, D, and T represent the set of generator nodes, the set of load nodes, and the set of scheduling time periods, respectively; L is the set of lines; P g,d,t P represents the energy exchange volume between generator node g and load node d at time t. g,t P represents the active power at generator node g at time t. d,t P represents the active power of load node d at time t. l,t The actual power flow of transmission line l at time t; x represents the energy exchange between generator node g and load node d at time t on transmission line l. ij X is the impedance value of branch l. ig This represents the element in the i-th row and g-th column of the power grid impedance matrix X. jg X id X jd All of these are elements in the power grid impedance matrix X.

3. The power grid transmission section identification method based on the improved PageRank algorithm and network transformation according to claim 1, characterized in that, Step 2 is based on the energy transaction volume P between the generator node and the load node. g,d,t The virtual circulation is set up and the Google matrix G in the original PageRank algorithm is corrected, as follows: Step 2.1: Add a virtual power flow pointer to g from load node d, with the power flow size being the energy trade volume P between generator node and load node. g,d,t ; Step 2.2: Set up virtual current conversion: Assume that the generator sets on a certain generator node transmit active power flow to multiple load nodes through transmission lines; based on the energy transaction data between generator nodes and load nodes calculated by the power system optimization operation model, the active power transmission path between generator nodes and load nodes is redirected from the load nodes to the generator node, and the weight of the virtual circulation is set according to the specific transaction volume. Then, all active power sent out by the generator returns to the generator node in the form of virtual circulation. By setting up a virtual circulation, the Google matrix G in the PageRank calculation formula is modified. The modified Google matrix is ​​denoted as G. * The specific formula is as follows: In equation (6), P ij For the active power flow of transmission line ij, P out (j) represents the outflow power of node j when considering virtual circulation.

4. The power grid transmission section identification method based on the improved PageRank algorithm and network transformation according to claim 3, characterized in that, Step 3 specifically includes: Step 3.1: Correct the restart vector e in the PageRank algorithm, denoted as e* after correction, as follows: Where C(i) is the network constraint coefficient of node i, which is determined by the topology of the power grid; n is the number of nodes in the power grid; Step 3.2: Calculate the network constraint coefficients C(i) for node i: Where q represents the common neighbor nodes of nodes i and j, and p ij p represents the proportion of node j to all adjacent nodes of node i. iq p represents the proportion of node q to all adjacent nodes of node i. qj Let Γ(i) represent the proportion of node j to all adjacent nodes of node q; Γ(i) is the set of adjacent nodes of node i; B(j) and B(u) are the three-level degrees of node j and node u, respectively; B(i) and k(i) are the degrees of node i considering three-level and single-level neighbor nodes, respectively; x ij Let be the reactance value of line l; w be the neighboring nodes of node i, v be the neighboring nodes of node w; Γ(w) be the set of adjacent nodes of node w; k(v) be the degree of a single layer of node v; a ij Let be the connection weight between nodes i and j; N is the set of all nodes in the network. Step 3.3: The improved PageRank algorithm calculation formula is as follows: PR k+1 =δ G * PR k +(1-δ) e * (13) Where δ is the drag coefficient, PR k+1 and PR k These are the node criticality values ​​calculated in the (k+1)th and kth iterations, respectively.

5. The power grid transmission section identification method based on the improved PageRank algorithm and network transformation according to claim 1, characterized in that, The specific method for network conversion in step 6 is as follows: nodes and edges in the power system topology graph are interchanged, that is, transmission lines and buses are converted to each other to obtain a new topology graph. Simultaneously, the criticality index (PR) of the nodes in the original topology graph is adjusted. k The reciprocal of the result is used as the converted line weight.

6. The power grid transmission section identification method based on the improved PageRank algorithm and network transformation according to claim 1, characterized in that, Step 7 specifically involves: Step 7.1: Based on the energy trading volume between nodes, determine the range between generator nodes and load nodes where the energy trading volume is greater than the set threshold as the range for the transmission section search; Step 7.2: Within the search range of the transmission section, obtain the shortest path using the shortest path search algorithm.

Images